A New Yorker “Annals of Psychology” article by the columnist-pundit David Brooks, entitled “Social Animal”, appeared in the Jan. 17, 2011 issue of that magazine. I have to say that it was a fairly feeble article, apparently drawn largely from a Psych 101 textbook of several years ago. However, Brooks” article contained a little discussion of a topic that deserves careful consideration-- whether unborn babies learn from the voices that penetrate the uterus and come to their ears. Brooks stated: “Even when he was in the womb, [the infant] was listening for his mother’s voice, and being molded by it. French babies cry differently from babies who’ve heard German in the womb, because they’ve absorbed French intonations before birth. Fetuses who have been read “The Cat in the Hat” while in the womb suck rhythmically when they hear it again after birth, because they recognize the rhythm of the poetry.”
I certainly do not want to say that none of this is true, but I believe that the important implications of such early learning make it essential for us to examine with care the research behind the claims Brooks is making. The research report stating that French and German babies cry differently is by B. Mampe, A.D. Friederici, A. Christophe, and K. Wermke (“Newborns’ cry melody is shaped by their native language”),and was published in Current Biology in 2009 (Vol.19, pp. 1994-1997). Mampe and her colleagues recorded cries from 60 babies 2-5 days of age (half born of French-speaking mothers, half born of German-speaking mothers) while they were interacting with their own mothers-- having their diapers changed, preparing for feeding, or being soothed when spontaneously fussing. The cries were not the characteristic high-pitched pain cry of the young infant and were not recorded at times when there was a clear reason for pain like a heel-stick for a blood sample.
Mampe and her colleagues analyzed the ways the sound frequencies of these cries changed over the brief period involved in one exhalation of breath. Like adults, infants change their sound production rapidly in ways that can produce a rising tone or a falling one. The researchers were able to analyze each cry and create a graph indicating its pattern, and later to statistically compare all the babies of French mothers and all the babies of German mothers on their cry “shapes”, whether rising or falling. This was of interest because sound production by French speakers involves more rising tones than do the sounds produced by German speakers.
Because there is no reason to suppose that French people have different vocal anatomy than German people, it appears that those different sound patterns must be learned through experience. But when does that experience begin to have an effect? Mampe and her colleagues suggest that the babies must have learned prenatally to use the sound pattern of their mother’s language.
Of course this is a possibility. However, there are some details about this study that need to be examined before rushing to a conclusion. Mampe’s group reported statistical comparisons between the French-background group and the German-background group’s cries, but it’s not quite clear how they did all the steps in their analysis. When there is more than one cry from each baby, and not necessarily the same number of cries from each, there are particular ways to do a statistical analysis; most statistical tests are designed to work on one score or reading from each participant. The Mampe paper does not make it clear how this problem was solved. In addition, the graphic display of the results shows a very considerable overlap between the sound patterns produced by the two groups, with some “German-listening” babies crying with the French-like rising tones. Whatever was happening here, it would seem that whether or not hearing experience made a difference, there were some unidentified factors at work too.
There is a more obvious issue as well, having to do with the basic research design. When the babies’ cries were recorded, they were interacting in some way with their own mothers. Each French-experienced baby was with a French-speaking mother, and each German-experienced baby was with a German-speaking mother. Whether the baby continued to cry or stopped depended in part on the mother’s response. It’s also possible that how the baby managed its rising or falling tone also depended on the mother’s response. Possibly (but of course we can’t tell on the basis of Mampe’s research), a French-speaking mother responded to “French-like” cries, with their rising tone, as if she recognized the sound as language, and to “German-like” cries, with their falling tone, as if this was a vocalization but not related to speech. German-speaking mothers would reverse this, but in both cases the mothers might become more attentive, change facial expression, and speak when they heard the sounds that resembled their own speech, thus rewarding and prolonging a particular sound pattern. To really know about this we would need to switch German-speaking mothers to briefly caring for French babies, and French-speaking mothers to briefly caring for German babies. We might find out that because of her interest in understanding a baby’s intentions, a mother does things that shape even very early sound production.
Let me switch back for a minute to Brooks’ New Yorker article and that final claim that babies suck rhythmically because they recognize the rhythm of the poetry. This statement is presumably based on the well-known study by DeCaspar and Spence in which unborn babies were allowed to hear either readings of the Dr. Seuss classic “The Cat in the Hat” or an alternative with simple but different rhymes. After they were born, the babies were given a nipple device, like a pacifier. If they sucked in one way (long rests between sucks) they would hear one of the readings; if they sucked another way (short rests between sucks), they would hear the other. The babies tended to suck in the way that would let them hear the familiar verse. But… it did not matter whether the person reading the verses was the baby’s mother or someone else. They sucked in a particular pattern in order to hear the familiar verse, not to hear their mother’s voice. And, contrary to what Brooks implies, the babies did not imitate the rhythm of the verse with their sucking rhythms; they just used sucking as a way to choose what they listened to.
The moral of this story, I suppose, is that year by year we learn more about how young humans develop. It’s not easy to find this out-- we have to discover clever ways to get babies to tell us about themselves. A study that seems very definite in its results can suddenly be seen to need substantial follow-up. Regrettably, much of what’s written about children (like Brooks’ New Yorker article) does not include the essential thinking that’s needed before conclusions are drawn. Caveat lector!
When I was a small girl, the question of "Can you cry in [this language]" was fascinating to me.
ReplyDeleteThen of course I read about sound production and mother's responses, which do seem to be the key factors here.
The poetry claim may or may not sound out there. And there are multiple factors there, too, at work. It's interesting to see the agency of the sucking babies!
(And I worked this way trying to teach dogs to sing or respond to me when I was singing).
Have also been reading the comments of Kit Whitfield, in which mothers observe how they are getting to know their babies as people and not only as need machines.
I think the article you might be referring to is THE COMPOSURE CLASS:
THE COMPOSURE CLASS: SOCIAL ANIMAL by David Brooks [17.1.2011]
and there is a definite funny, sardonic tone in there. Which might make it all the more dangerous.
Although Brooks uses the term "composure class" in the text, the title is just "Social Animal" in both the print version and the on line version you give the link for. Don't know why!
ReplyDeleteI must look at the Kit Whitfield material-- thanks.